Why Most VC Research is Useless (And How to Fix It)

Most venture capital research is performance art.

That is a hard sentence to write as someone whose entire company is built around research infrastructure. But I will defend it. The majority of what passes for research inside VC firms, family offices, and angel portfolios is not actually helping anyone make better decisions. It is filling time, generating memos that no one rereads, and producing the illusion of rigor around gut-driven investment calls.

I say this with enough firsthand evidence that I stopped being polite about it a year ago. And I have strong opinions about what real research looks like, because Brevoir exists to provide it.

This post is the diagnosis. What is broken, why it is broken, and what the fix actually looks like.

The Research Output Nobody Reads

Walk into any VC firm and look at the research artifacts it produces. You will find something like this:

• Weekly sector digests, written by associates, consumed for about 90 seconds before being filed.
• Individual investment memos, 10 to 40 pages, read carefully once during IC and never opened again.
• Quarterly market reports, distilled from the same five newsletters everyone else reads.
• One-off thematic deep dives, triggered by whatever topic the partnership got interested in three months ago.

The volume looks impressive. The shelf life of the individual artifacts is measured in days.

Here is the uncomfortable question: how many investment decisions in the last year were actually made differently because of any of this output? Not "informed by." Actually changed. If the honest answer is "maybe one or two," the entire research stack is underperforming.

Why It Stays Broken

If this is so widely known, why does it persist? Because incentives inside VC firms reward research production, not research quality.

Research Is Performed for the Audience, Not the Decision

Associates write memos because memos are their evidence of work. Partners read memos selectively because the memo exists as process, not because it changes minds. The document is a performance of rigor. It signals that due diligence has been done, so that if the deal goes bad, there is evidence the team "did the work."

This is the dominant failure mode. Research becomes compliance, not epistemology.

The Research Is Too Slow

By the time a sector report is written, edited, and distributed, the market has moved. A six-week research sprint in a hot sector produces a document that is out of date by the time it lands. In private markets, where round dynamics move in days and weeks, monthly or quarterly research cycles are structurally too slow.

Newsletter-driven investing→ is just the externalized version of this problem. Same staleness, different author.

The Research Is Not Personalized

The same sector report is sent to every partner in the firm, even though different partners have different theses, different geographies, different stages, and different conviction criteria. General research for specialized investors is almost never actionable at the point of decision.

The Research Has No Freshness Mechanism

Once a memo exists, it is frozen. The startup it describes keeps evolving. Competitors ship new products. New funding rounds reshape the competitive landscape. The memo captures a moment and pretends it is a landscape.

Three months after a memo is written, it is typically wrong in several material ways. Nobody updates it. The firm operates on stale intelligence and calls it research.

The Four Problems Research Must Solve

Before describing what works, let me describe what research actually needs to do to earn its place in an investment process.

1. Inform Decisions at the Point They Are Made

Research is only useful if it changes or sharpens a decision. That means it has to be accessible at the moment the decision is being made, in a form the decision maker can use, without a week of lead time.

If a partner in a pipeline meeting needs to decide whether to pursue a deal, the relevant intelligence has to be available in minutes, not weeks. If an IC is debating a competitive company's trajectory, real-time data beats a three-month-old memo.

2. Be Verifiable

Every claim in a piece of research should be traceable to a source. "We believe fintech infrastructure is heating up" is not research. "Funding in fintech infrastructure startups grew 73% year-over-year in Q1 2026, based on 38 disclosed rounds totaling $2.1B" is research. The first is an opinion. The second is a statement that can be challenged, updated, and used.

Research without source-level verification is just narrative in a business-school font.

3. Be Continuously Updated

The world does not pause between memo publications. Real research is a living system, continuously refreshed as new data arrives. Static PDFs are not research. They are snapshots.

4. Be Personalized to the Thesis

A generalist investor's research needs are different from a climate tech specialist's research needs, which are different from an emerging markets investor's research needs. One-size-fits-all research is one-size-fits-no-one at the point of decision.

Research infrastructure has to know your thesis, your portfolio, your target sectors and stages, and deliver intelligence filtered through that specific lens.

What Real Research Looks Like

Real research, at the level of infrastructure a serious private market investor actually needs, looks very different from the memo-and-newsletter stack most firms still run on.

Real-Time Sector Monitoring

Instead of quarterly sector reports, continuous sector monitoring. Funding velocity, round sizes, investor concentration, exit multiples, and narrative shifts, updated as the data comes in, with alerts when material changes happen.

The question "what is happening in climate tech infrastructure right now" should have an answer in 15 seconds, not require a three-week research sprint.

Structured Competitive Intelligence

For every portfolio company and every serious prospect, an always-current competitive landscape. Who just raised. Who just hired executives from a key competitor. Which products shipped. Which partnerships announced. What the regulatory backdrop looks like.

A competitive landscape that is static for more than two weeks is already wrong. Real research keeps this view continuously fresh.

Thesis-Matched Signal Detection

Signals filtered through your specific investment thesis. If you invest in B2B vertical SaaS at seed stage in North America and Europe, you should see signals about those companies, in those sectors, at that stage, in those geographies. Not a general market roundup.

This is the fundamental difference between personalized intelligence and generic research. Personalization is not a feature. It is the core of the product.

Source-Verified Data, With Attribution

Every data point in the intelligence layer should have a source attached. A claim about a company's revenue should link to the article, filing, or interview where that claim originated. A funding round should link to the announcement. A hire should link to the LinkedIn post or press release.

Source attribution is not a luxury. It is the foundation of trust in the data. Without it, research is just confident assertion.

Leading Indicators, Not Lagging Summaries

By the time something shows up in a quarterly report, it is already known by everyone. Real research tracks leading indicators: hiring surges before they are announced as growth, product launches before they are covered by the press, funding signals before rounds close, regulatory shifts before they are interpreted as sector tailwinds.

These signals are where research can still generate alpha. Lagging summaries cannot.

Who Should Do the Work

Another broken assumption in most firms: that humans should be doing this work.

For the last 15 years, the default answer to "we need more research" has been "hire another associate." Associates write memos. Associates maintain market maps. Associates track portfolio news.

That default is breaking down. The honest truth is that most of the research work inside VC firms is now better done by software. Not because humans are bad at research, but because the specific tasks that dominate research workloads (data gathering, cross-referencing, continuous monitoring, structured synthesis) are exactly the tasks that modern AI and dedicated intelligence platforms handle well.

The human work that remains is higher-leverage: judgment calls, relationship-driven intelligence, thesis development, contrarian interpretations of the data. But the raw research infrastructure should be machine-powered.

AI in VC→ is not a hypothetical. It is already reshaping how research actually gets done at the firms that take it seriously.

What Firms Should Do

If you are running a fund, angel portfolio, or family office, and recognize your research stack in the "broken" description above, here is a pragmatic path forward.

1. Stop Producing Research Nobody Reads

Audit your research output over the last year. Which artifacts were read more than twice, by more than two people? Kill or drastically reduce everything else.

Reallocate the time saved toward decision-adjacent research: investigation in support of live deals, not general market coverage.

2. Invest in Real-Time Intelligence Infrastructure

A dedicated intelligence platform, whether Brevoir or an internally built equivalent, is now a baseline requirement for competitive firms. Real-time sector data, competitive monitoring, and thesis-matched signals are the new foundation.

Firms that try to do this with spreadsheets, Google alerts, and junior staff manually tracking companies will be structurally slower than firms with dedicated infrastructure. Over a portfolio, that speed gap is worth substantial performance.

3. Make Research Continuous, Not Episodic

Stop running research in quarterly sprints. Move to a continuous-update model. Every portfolio company view should be always-current. Every sector dashboard should refresh automatically. Every competitive landscape should update as new data arrives.

Episodic research produces artifacts. Continuous research produces situational awareness.

4. Redefine the Researcher's Role

The researcher of 2026 is not a memo writer. They are an analyst who configures the intelligence infrastructure, interprets the signals it surfaces, pushes back on the automated conclusions with human judgment, and maintains the connection between thesis and data.

That is a materially higher-value role than "associate who writes sector reports." It also attracts materially better talent, because it is actually interesting work.

The Broader Pattern

This is not a VC-specific problem. It is a symptom of a broader shift that has already happened in public markets and is finally happening in private markets.

Public market research infrastructure (Bloomberg, FactSet, Refinitiv, S&P Capital IQ) exists because public market investors decided 30 years ago that institutional-grade, continuously updated, source-verified data infrastructure was a non-negotiable cost of doing business. No serious public market firm runs on newsletter summaries and bespoke memos.

Private markets have lagged this by decades. Most private market firms are still running on 1995 research workflows (individual memos, scattered newsletters, ad hoc phone calls) in 2026. The gap is closing, but slowly, and the firms that close it first will have an enormous advantage over the ones that do not.

Private markets need their own Bloomberg.→ This is what that actually means in practice. Not a data terminal for the sake of aesthetics. An intelligence infrastructure that makes real research possible at the pace the market actually moves at.

The Hard Truth

Most of what firms currently call "research" is not actually research. It is documentation. It exists to demonstrate rigor, not to produce insight. The investors who recognize this and rebuild their research function around real-time, source-verified, thesis-matched intelligence are not going to produce slightly better memos. They are going to compound at higher IRRs, because their research is actually influencing decisions rather than decorating them.

This is one of the handful of infrastructure upgrades in private markets with asymmetric upside. The cost of getting it right is modest. The cost of staying on the old model, while competitors upgrade, compounds against you for years.

That is the case we make for Brevoir every day, and it is not a marketing line. It is a genuine belief that private market intelligence infrastructure is currently where public market infrastructure was in the 1980s, and the firms that adopt early will win.

If your research stack looks like newsletters, PDFs, and associates, it is time to upgrade. Real-time sector intelligence, thesis-matched signals, source-verified data, and continuous competitive monitoring are what modern research looks like. See what that infrastructure actually looks like→, and decide whether your current research stack is still worth the time it consumes.

Written by

Nabil Abuhadba

CEO and founder of Brevoir. Building the intelligence infrastructure for private markets. Previously obsessing over data, startups, and the future of investing.

@nabuhadLinkedIn